Use --permission-mode to Start Claude in a Specific Permission Mode
Instead of toggling permission modes after launch with Shift+Tab, you can start Claude Code in exactly the mode you want from the command line.
claude --permission-mode plan
This launches Claude directly in Plan mode, where it researches and reasons without making changes. Other available modes include default (the normal ask-before-acting mode) and auto (which requires --enable-auto-mode and a Team plan).
This is especially useful when you want to enforce a specific workflow from the start. For example, you might want to always begin a new feature in Plan mode so Claude maps out the approach before touching any files:
# Start a planning session for a new feature
claude --permission-mode plan "Design the authentication flow for our API"
# Or combine with other flags for CI
claude --permission-mode plan --print "Analyse this codebase and suggest improvements"
You can also compose it with --allow-dangerously-skip-permissions for advanced CI setups where you want plan mode as the default but need the option to escalate:
claude --permission-mode plan --allow-dangerously-skip-permissions
Start every session with the right guardrails, not just the ones you remember to toggle.
Log in to leave a comment.
When Claude writes error messages, button labels, validation text, or onboarding flows, it defaults to generic developer-speak. Add a "Users" section to your CLAUDE.md describing who your actual users are — their technical level, industry jargon, and what they care about — so Claude writes copy that makes sense to THEM, not to developers.
Use the --agent flag with custom markdown files in .claude/agents/ to launch purpose-built Claude sessions with restricted tools and scoped system prompts.
Every project has traps — the billing module that silently fails if you forget to queue the job, the legacy table with column names that don't match the model, the config value that must be set before tests run. Document these gotchas in your CLAUDE.md so Claude avoids the same mistakes your team spent days debugging.